AI Could Make Our Writing More Alike. And That May Be the Bigger Creative Risk

AI & The Author

Quick Shortcuts: The story | Why writers should care | The danger to voice | Where AI still helps | What authors can do | Discussion

AI Could Make Our Writing More Alike. And That May Be the Bigger Creative Risk

The argument about artificial intelligence and writing has spent a lot of time on ownership: who wrote the words, whose books trained the model, and whether AI-generated work should be disclosed. A fresh piece of research raises a quieter question that may matter just as much to working authors: what happens to human writing when millions of us begin consulting the same kinds of machines while we create it?

The Bookseller reported on 16 September that researchers are warning widespread use of large language models could homogenise human writing and thought. The concern is not simply that AI can produce generic prose. It is that repeated interaction with systems trained to predict statistically likely language may subtly pull human writers towards similar vocabulary, structures, arguments and perspectives.

For authors, that lands uncomfortably close to the heart of what makes a book worth reading in the first place. Readers do not need another technically competent sentence. They need a voice. They need the odd choice, the personal rhythm, the unexpected observation and sometimes the sentence an algorithm would never have predicted.

This is different from the usual AI argument

There is an important distinction here. I do not believe every use of AI around writing is equivalent to asking a machine to write a novel. I write my books. AI can be useful around that work: research support, testing an idea, organising material, visualising a world, helping with marketing, or handling some of the repetitive business surrounding an author's life.

That is AI assistance. It is not the same thing as outsourcing the creative decisions that make the work yours.

But the homogenisation argument introduces a grey area. A writer may still type every final word and yet repeatedly ask an AI system how a scene should work, how a character should respond, what the next plot beat should be, or how a paragraph should be rewritten. At some point the danger is not that the machine becomes the named author. It is that the writer's own instincts become progressively less visible.

The value of the strange sentence

Good fiction is full of inefficiencies. Characters make poor decisions. Authors linger where another writer would hurry. A metaphor may be peculiar but memorable. A scene may violate the supposedly optimal structure and work precisely because it does.

Generative AI is extraordinarily good at recognising patterns. That is one of its strengths. Creative writing, however, often becomes interesting when somebody deliberately or instinctively breaks one.

If thousands of writers ask similar models for developmental advice, plot solutions and stylistic improvements, the question is whether the resulting books begin to converge—not because they were generated by AI, but because their human authors were repeatedly nudged towards the same statistically persuasive answers.

Writers are already drawing the boundary

This concern arrives alongside a much more visible argument about professional authorship. British screenwriter and Writers' Guild of Great Britain president Jack Thorne recently warned that some professional writers are using AI to generate entire scripts while presenting the work as their own. His criticism is aimed at undisclosed AI authorship, but the two debates meet at the same point: writers need to decide which parts of the creative process they are prepared to surrender.

The answer does not have to be zero technology. Modern writing is already surrounded by it. Word processors correct spelling, software checks grammar, search engines support research, and publishing systems increasingly automate administrative work. The meaningful question is whether the technology helps the writer express an idea or begins supplying the idea, judgement and voice itself.

A practical rule for authors

My own rule is becoming increasingly simple: use AI most confidently at the edges of the creative work, and become progressively more cautious as it approaches the centre.

Research assistance? Useful. Visualising a starship or character while worldbuilding? Useful. Organising campaign ideas, analysing marketing options or repurposing material I have already written? Useful. Asking a machine to decide what my character feels, write the emotional climax or manufacture the finished chapter? That crosses into territory I want to remain mine.

There is another practical test. Before accepting an AI suggestion, ask whether you would have arrived at it yourself. If the answer is no, do not automatically reject it—but interrogate it. Why is it better? Does it sound like you? Does it make the story more distinctive, or merely more conventional?

The objective should not be to prove that no machine ever touched the creative process. It should be to ensure that when somebody finishes the book, the voice they encountered was unmistakably the author's.

Why this matters now

AI-generated material is becoming easier and cheaper to produce, while online marketplaces already make discoverability difficult for human writers. In that environment, distinctiveness becomes more valuable, not less. Trying to make human writing behave more like machine writing may therefore be exactly the wrong competitive response.

The opportunity for authors is the opposite: use technology to remove friction around the work while protecting the messy, personal, unpredictable centre of it.

Because if AI eventually becomes capable of producing perfectly competent prose on demand, competence will not be the scarce commodity. Human perspective will.

The question

If AI can make writing cleaner, faster and more structurally predictable, but potentially more alike. How much imperfection would you deliberately keep in order to preserve your own voice?

Sources & further reading

The Bookseller, 16 September 2026: Researchers warn use of AI in writing could ‘homogenise human writing and thought’.

The Guardian, 14 September 2026: Jack Thorne warns some fellow scriptwriters are using AI ‘to cheat’

About Rob Frankson

Rob Frankson is a science-fiction author and creator of the Near Galaxy Saga. Through 121 Minutes he writes about storytelling, publishing, creativity and the changing relationship between authors and artificial intelligence.

Want more?

Join 121 Minutes

For more writing, publishing, science fiction and AI & The Author, visit 121minutes.com and Join the 121 Minutes community for the latest AI & The Author briefings, writing updates and new fiction.

AI & Editorial Transparency

AI & The Author is edited and published by Rob Frankson. Artificial intelligence is used to assist with news research, initial drafting, content organisation and supporting imagery. All articles are reviewed and, where necessary, edited by Rob Frankson before publication. The opinions, editorial position and final decision to publish remain the author's.